I have a media server that has changed from "Active for Tape and Disk" to simply "Active for Disk". Jobs that get scheduled for it stall on "waiting for media" (even if the media is disk).
A reboot of the media server didn't magically clear anything up. I have seen something similar before where the media server has some media assigned to it in the EMM, even though the media is no longer there.
Has anyone seen this condition before and have a suggestion? I am unable to find my notes on how I was able to query the EMM to find what volumes were assigned to this server so that I can cancel the assignment.
NBU is 18.104.22.168.
TECH69625 walks though the processes for "How to troubleshoot and correct problems with media servers when the status changes in the Media Servers area of the NetBackup Administration GUI."
But first off , if you can, try bouncing the netbackup services on the media server. You can also remove and readd the robot and drives via the device wizard. That might do it.
That particular article appears to focus on the network connectivity (and the name resolution associated) for media servers. This particular media server hasn't had any changes, and I've confirmed that all the names of the interfaces are being resolved on both the master and the media server as expected.
I was unable to completely bounce the daemons on the media server initially. That was why I rebooted it to be sure it was clean. Before the reboot I had bptm processes that would not die. I've confirmed since the reboot that I can see all the tape drives with "scan" and they appear in the windows device manager.
Have you checked and verified that required licenses are installed and working
Enterprise Server/Client license
Library Based Tape drives
If shared drives, SSO.
Check for all drive paths are all correct ?8
If all is well, then execute below command
nbemmcmd -updatehost -machinename <media_server> -machinestateop set_tape_active -machinetype media -masterserver <master_server>
The symptoms sound like comms issues between master and this media server.
Are Master and Media server both Windows servers?
All on same NBU patch levels?
Any one of them with more than one NIC (multiple IPs/hostnames)?
This TN http://www.symantec.com/docs/TECH70827 explains how EMM heartbeats are sent every 5 minutes from the media server. Lack of heartbeats from media server will cause master server to change state.
Best to set up logging as per TECH70827 to trace heartbeats initiated by media server and received by master.
About device/media allocations - you are probably looking for nbrbutil command:
Find MDS allocations with:
nbrbutil -releaseMDS <mdsAlocationKey>
nbrbutil -resetMediaServer <mediaserver>
I came across similar issue multiple times... not sure whether its kind -o-bug or defect to sy..
Simply remove all the STU for that media server and recreate it (especially tape based STU)
Just verify STU has correct density for robots and drives as same must match with your medias else backups won't work.
use nbemmcmd for deleting all associated STU for affected media server (if DISK BASED STU is created do not use nbemmcmd) instead manually remove it from GUI
Give it a try.. This generally does work like charm...